Showing math.OCShow all
2 papers · 1 filter
math.OC2024
Stochastic Newton Proximal Extragradient Method
Ruichen Jiang, MichaÅ DereziÅski, Aryan Mokhtari
Stochastic second-order methods achieve fast local convergence in strongly convex optimization by using noisy Hessian estimates to precondition the gradient. However, these methods…
math.OC2024
Second-order Information Promotes Mini-Batch Robustness in Variance-Reduced Gradients
Sachin Garg, Albert S. Berahas, MichaÅ DereziÅski
We show that, for finite-sum minimization problems, incorporating partial second-order information of the objective function can dramatically improve the robustness to mini-batch s…